Kaltura Model Context Protocol (MCP) Server

Name: Kaltura Model Context Protocol (MCP) Server
Author: zoharbabin

@zoharbabin

Repository Homepage

About Kaltura Model Context Protocol (MCP) Server

Kaltura-MCP server powering agentic AI workflows with Kaltura's API

Basic information

Config

Add this server to your MCP-compatible client using the configuration below.

{
  "mcpServers": {
    "kaltura-mcp": {
      "command": "python",
      "args": [
        "setup_env.py"
      ]
    }
  }
}

Tools

entry_id (required)

search_text, limit

General analytics data for reporting and analysis

Time-series data optimized for charts

Detailed viewer retention analysis throughout videos

Live analytics updated every ~30 seconds

Quality of Experience (QoE) and streaming performance

Location-based analytics at various granularities

Discover all available analytics functions

entry_id (required), flavor_id

entry_id (required), width, height, second

query (required), search_type, match_type, specific_field, boolean_operator, include_highlights, custom_metadata, date_range, max_results, sort_field, sort_order

entry_id (required)

caption_asset_id (required)

entry_id (required)

Overview

What is Kaltura Model Context Protocol (MCP) Server?

A Model Context Protocol (MCP) server that provides secure, read-only tools for managing Kaltura API operations. It enables AI assistants to search, discover, and analyze Kaltura media content safely through a defined set of tools and resources.

How to use Kaltura Model Context Protocol (MCP) Server?

Install the package with pip install kaltura-mcp, then either use the interactive setup_env.py script or manually create a .env file with your Kaltura credentials. For local use, configure Claude Desktop to point to the installed command. For remote hosting, deploy the server (e.g., kaltura-mcp-remote) and configure users via OAuth/JWT. The server supports both stdio (local) and HTTP/SSE (remote) transports.

Key features of Kaltura Model Context Protocol (MCP) Server

Media Discovery: Search and browse media entries with advanced filtering
Content Access: View captions, transcripts, and attachment content
Category Management: Browse and explore content categories
Analytics Suite: Retrieve viewing analytics, retention, real-time metrics, and geographic breakdowns
Read-Only Operations: All tools are safely limited to read-only access
Multi‑mode Deployment: Works locally with Claude Desktop or as a hosted remote service

Use cases of Kaltura Model Context Protocol (MCP) Server

Search for recent videos in your Kaltura account using natural language
List and browse content categories to organize media
Retrieve detailed analytics reports on video performance and viewer engagement
Generate accessibility audits of captions and attachments across your content library
Obtain download URLs and thumbnail previews for media entries

FAQ from Kaltura Model Context Protocol (MCP) Server

What type of access does this server provide?

The server enforces read-only operations. It does not support creating, updating, or deleting any Kaltura data.

How do I obtain the required Kaltura credentials?

You need a Service URL, Partner ID, Admin Secret, and User ID from your Kaltura account. These can be found in KMC under Settings → Integration Settings.

What are the two deployment modes?

Local (stdio) mode directly connects to Claude Desktop; remote (HTTP/SSE) mode runs as a hosted server with OAuth/JWT authentication and multi‑tenant support.

Why do I get a “missing environment variables” error?

The .env file is missing or has incorrect permissions. Ensure it exists in the project directory and has permissions set to 600. You can run setup_env.py to recreate it.

How does authentication work for the remote server?

The remote server uses JWT tokens signed with a secret key. Users authenticate via an OAuth web flow, receiving a token they then include in their MCP client configuration.